Utilizing Machine Learning to Mitigate the Risk of Transfusion-Related Hemolytic Reactions

Overview

The monocyte monolayer assay is a cellular assay, an in -vitro procedure that mimics extravascular hemolysis. The assay is used to predict the clinical significance of red blood cell antibodies in transfusion candidates with intent to determine whether the patient needs to receive the expensive, rare, antigen - negative blood to avoid an acute hemolytic transfusion reaction that could lead to death. The assay requires a highly trained technician to spend several hours over a microscope, evaluating a minimum of 3,200 monocytes on a glass slide in a cumbersome process of repetitive counting. I employed machine learning to automate the identification and categorization of monocytes in slide images, presenting a significant improvement over the manual counting approach. The trained model can locate, identify, and categorize monocytes, separating them from the background and noise on the images acquired by an optical microscope camera. In the absence of a publicly accessible database containing these slide images for training the model, I acquired them at LifeShare Blood Center in Shreveport, Louisiana, and established an extensive public repository with the goal of serving as a comprehensive resource for automated analysis of monocytes. Utilizing the trained model I implemented on a Raspberry Pi, blood bank technicians can optimize their monocyte monolayer workflow, resulting in time and effort savings and ultimately contributing to expedited and improved medical diagnoses. Performance ev aluations demonstrate this approach can ease and accelerate the medical laboratory technician’s repetitive, cumbersome, and error -prone counting process, and therefore contribute to the accuracy of diagnosis systems. Maryland

Competition history

• JSHS 2024 Category not listed

Resources

• National JSHS abstract book ↗

Related projects

Closest projects by meaning, across every fair and year in the corpus.